Photodynamic therapy (PDT) is a non-surgical treatment of tumors in which non-toxic drugs and non-hazardous photosensitizing irradiation are combined to generate cytotoxic reactive oxygen species in situ. This technique is more selective than the commonly used tumor chemotherapy and radiotherapy.
PDT of tumors involves the combination of administered photosensitizer and local light delivery, both innocuous agents by themselves, but in the presence of molecular oxygen they are capable of producing cytotoxic reactive oxygen species (ROS) that can inactivate cells. Being a binary treatment modality, PDT allows for greater specificity and has the potential of being more selective, yet not less destructive, when compared with commonly used chemotherapy or radiotherapy (Dougherty et al., 1998; Bonnett et al., 1999; Kessel and Dougherty, 1999; Mazon, 1999; Hahn and Glatstein, 1999).
Porphyrins have been employed as the primary photosensitizing agents in clinics. Optimal tissue penetration by light apparently occurs between 650-800 nm. Porfimer sodium (Photofrin®, a trademark of Axcan Pharma Inc.), the world's first approved photodynamic therapy agent, which is obtained from hematoporphyrin-IX by treatment with acids and has received FDA approval for treatment of esophageal and endobronchial non-small cell lung cancers, is a complex and inseparable mixture of monomers, dimers, and higher oligomers.
Large amounts of work have been devoted to the synthesis of single pure compounds—so-called “second generation” sensitizers—which absorb at long wavelength, have well established structures and exhibit better differentiation between their retention in tumor cells and their retention in skin or other normal tissues. In order to optimize the performance of the porphyrin drugs in therapeutics and diagnostics, several porphyrin derivatives have been proposed in which, for example, there is a central metal atom (other than Mg) complexed to the four pyrrole rings, and/or the peripheral substituents of the pyrrole rings are modified and/or the macrocycle is dihydrogenated to chlorophyll derivatives (chlorins) or tetrahydrogenated to bacteriochlorophyll derivatives (bacteriochlorins).
Due to their intense absorption in favorable spectral regions (650-850 nm) and their ready degradation after treatment, chlorophyll (Chl) and bacteriochlorophyll (BChl) derivatives have been identified as excellent sensitizers for PDT of tumors and to have superior properties in comparison to porphyrins. Bacteriochlorophylls are of potential advantage compared to the chlorophylls because they show intense near-infrared bands, i.e., at considerably longer wavelengths than chlorophyll derivatives.
Tumor Vascular Targeting
Targeting photodynamic reagents for destruction of the tumor vasculature, as opposed to the tumor cells themselves, may offer therapeutic advantages since tumor-cell growth and development critically depend on continuous oxygen and nutrient supply (Ruoslahti, 2002). Such vascular damage may include thrombus formation and further restrict tumor blood perfusion (Huang et al., 1997). Furthermore, targeting the tumor vascular endothelial cell (EC) layer is expected to circumvent the poor penetration of tumor stroma by the therapeutic macromolecules (Huang et al., 1997; Burrows and Thorpe 1994). Although tumor blood vessels might be affected by the tumor microenvironment and acquire a tumor associated “signature”, they are not malignant and less likely to develop drug resistance. Furthermore, when a targeted antivascular agent is also active against the tumor cells, additional gains in efficacy can be expected. Thus, by combining antivascular properties with antitumor cytotoxic activities in one drug, its efficacy can be expected to increase and, consequently, decrease the required effective cytotoxic dose. In addition to ECs, tumor cells have also been shown in one case to comprise part of the luminal surface mosaic of the tumor blood vessels (Ruoslahti, 2002; Chang at al, 2000). Consequently these tumor cells are thought to be directly exposed to the blood and freely interact with therapeutic macromolecules that otherwise are unable to cross the endothelial barrier.
Selective vascular targeting can rely on the differential susceptibility and consequent response to therapeutic agents of tumor and normal blood vessels. Alternatively, differential endocytosis may promote selective uptake of cytotoxic or other therapeutic agents. Recent studies have suggested organ/tissue specific properties for vascular ECs (Ruoslahti, 2002). The blood vessels in different tissues are likely to express tissue specific endothelial markers that are mostly unknown. Pathological processes such as inflammation, ischemia and malignancy can also impose their signature on the respective vasculature (Ruoslahti, 2002; Ruoslahti and Rajotte, 2000; Ruoslahti, 2000; Rajotte et al., 1998; Arap et al., 1998). The biochemical features that characterize blood vessels in tumors may include angiogenesis-related molecules such as certain integrins. The integrins αvβ3, αvβ5 and α5β1 have been identified in expression patterns typical for angiogenic vascular ECs associated with tumors, wounds, inflammatory tissue, and during vascular remodeling (Brooks et al, 1994a; Brooks et al, 1994b; Brooks et al, 1995; Elceiri and Cheresh, 1999). Endothelial-cell growth factor receptors, proteases, peptidases, cell surface proteoglycans and extracellular matrix (ECM) components have also been described (Ruoslahti, 2000). This rich repertoire of heterogenic molecules and processes may provide new opportunities for targeted delivery of therapies.
Different strategies have been pursued to achieve this goal. Circulating peptides, peptidomimetics or antibodies that target specific sites in the vasculature are attractive as carriers for therapeutics and diagnostic agents offering theoretical advantages over such conjugates that directly target tumor cells, mostly situated beyond physiological barriers such as the blood vessel wall.
Chaleix et al., 2003, disclose the synthesis of RGD-porphyrin conjugates as potential candidates for PDT application, in which the unmetalated porphyrin macrocycle is substituted at each of the positions 10,15,20 by 4-methylphenyl or acetylatedglucosyloxyphenyl and at position 5 by a residue of a linear RGD-containing peptide linked to the macrocycle via a spacer arm.
Selective uptake of RGD-containing peptides by endothelial and tumor cells via αvβ3 and αvβ5 integrins
The arginine-glycine-aspartic acid Arg-Gly-Asp (RGD) motif of ECM components, like fibronectin (Pierschbacher and Ruoslahti, 1984) and vitronectin, binds to integrins (Ruoslahti and Pierschbacher, 1987; D'Souza S E et al., 1991; Joshi et al, 1993; Koivunen et al., 1994). Integrin-mediated adhesion leads to intracellular signaling events that regulate cell survival, proliferation, and migration. Some 25 integrins are known, and at least eight of them bind the RGD motif as the primary recognition sequence in their ligands.
Data obtained by phage display methods (Pasqualini and Ruoslahti, 1996) screening for RGD-containing peptides, have shown their selective binding to endothelial lining of tumor blood vessels (Ruoslahti, 1996; Pasqualini et al., 1997). Because the expression of integrins is reported to be high on activated, but more restricted on quiescent, ECs, small synthetic RGD-containing peptides have been proposed as antagonists impairing the growth of vascular endothelial and tumor cells. RGD peptides also retard signal transmission, affect cell migration and induce tumor cell regression or apoptosis (Su et al., 2002). RGD-analogues are used in tumor imaging (Haubner et al., 2001), anti-angiogenesis approaches (Kawaguchi et al., 2001; Pasqualini et al., 2000), and in tumor targeting of radionucleotides (van Hagen et al., 2000) and chemotherapeutic drugs (Arap et al., 1998; Zitzmann et al., 2002).
Integrins are also expressed on cancer cells and therefore play an important role in the invasion, metastasis, proliferation and apoptosis of cancer cells. Metastasis invasion of tumor cells into preferred organs may represent cell-homing phenomena that depend on the adhesive interaction between the tumor cells and organ-specific endothelial markers (Ruoslahti and Rajotte, 2000). By binding to integrin of either endothelial or tumor cells, RGD peptides are capable of modulating in vivo cell traffic by inhibition of tumor cell-ECM and tumor cell-EC attachments, which are obligatory for metastatic processes. Several studies have indicated that RGD-containing compounds can interfere with tumor cell metastatic processes in vitro (Goligorsky et al., 1998; Romanov and Goligorsky 1999) and in vivo (Saiki et al., 1989; Hardan et al., 1993).
Peptides that are specific for individual integrins are of considerable interest and of possible medical significance. The αvβ3 integrin was the first integrin shown to be associated with tumor angiogenesis. RGD peptides that specifically block the αvβ3 integrin show promise as inhibitors of tumor and retinal angiogenesis, of osteoporosis and in targeting drugs to tumor vasculature (Assa-Munt et al., 2001). Coupling of the anticancer drug doxorubicin or a pro-apoptotic peptide to an αvβ3 integrin-binding RGD peptide yields compounds that are more active and less toxic than unmodified drugs when tested against xenograft tumors in mice (Ruoslahti, 2000; Arap et al., 1998; Arap et al., 2002; Ellerby et al., 1999).
U.S. Pat. No. 6,576,239, EP 0927045 B1 and WO 98/010795 (all of The Burnham Institute, Inventors: E. Ruoslahti and R. Pasqualini) disclose a conjugate comprising a tumor homing peptide comprising the amino acid sequence RGD or NGR, said tumor homing peptide linked to a therapeutic or diagnostic moiety, provided said moiety is not a phage particle. The therapeutic moiety may be a cytotoxic agent or a cancer chemotherapeutic agent such as doxorubicin. The conjugate selectively homes to angiogenic vasculature upon in vivo administration. The tumor homing peptide may be a peptide of up to 20 or 30 amino acids or of 50 to 100 amino acids in length, linear or cyclic. One preferred peptide is the cyclic nonapeptide, CDCRGDCFC or H-Cys*-Asp-Cys*-Arg-Gly-Asp-Cys*-Phe-Cys*—NH2.
Selective Vascular Response Induced in Tumors by Photodynamic Therapy (PD7)
Application of novel bacteriochlorophyll (Bchl) derivatives as sensitizers in PDT has been reported by our group in recent years in the scientific literature (Zilberstein et al., 2001; Schreiber et al., 2002; Gross et al., 1997; Zilberstein et al., 1997; Rosenbach-Belkin et al., 1996; Gross et al., 2003a; Koudinova et al., 2003; Preise et al., 2003; Gross et al., 2003b) and in the patent publications U.S. Pat. Nos. 5,726,169, 5,650,292, 5,955,585, 6,147,195, 6,740,637, 6,333,319, 6,569,846, 7,045,117, DE 41 21 876, EP 1 246 826, WO 2004/045492, WO 2005/120573. The spectra, photophysics, and photochemistry of Bchl derivatives have made them optimal light-harvesting molecules with clear advantages over other sensitizers presently used in PDT. These Bchl derivatives are mostly polar and remain in the circulation for a very short time with practically no extravasations into other tissues (Brandis et al., 2003). Therefore, these compounds are good candidates for vascular-targeted PDT that relies on short (5-10 min) temporal intravascular encounter with light and higher susceptibility of the tumor vessels to the PDT-generated cytotoxic ROS.
Recent studies performed by our group showed that primary photosensitization is intravascular with rapid development of ischemic occlusions and stasis within the illumination period. This process also induces photochemically induced lipid peroxidation (LPO) and early EC death that is primarily confined to the tumor vasculature (Gross et al., 2003a; Koudinova et al., 2003). Due to light independent progression of free radical chain reactions along with developing hypoxia, LPO and cell death spread beyond the vascular compartment to cover the entire tumor interstitium until complete necrosis of the tumor is attained around 24 hours post PDT. Hence, the primary action of PDT blocks blood supply and induces hypoxia that initiates, in a secondary manner, a series of molecular and pathophysiological events that culminate with tumor eradication.
Mitochondria, lysosomes, plasma membrane, and nuclei of cells have been evaluated as potential PDT targets. Since most PDT sensitizers do not accumulate in cell nuclei, PDT has a generally low potential of causing DNA damage, mutations, and carcinogenesis. Hydrophilic sensitizers are likely to be taken up by pinocytosis and/or endocytosis and therefore become localized in lysosomes or endosomes. Light exposure will then permeabilize the lysosomes so that sensitizers and hydrolytic enzymes are released into the cytosol (Dougherty et al., 1998).
PDT damage to plasma membrane can be observed within minutes after light exposure. This type of damage is manifested as swelling, shedding of vesicles containing plasma membrane marker enzymes, cytosolic enzymes and lysosomal enzymes, reduction of active transport, depolarization of plasma membrane, inhibition of the activities of plasma membrane enzymes, changes in intracellular Ca2+, up- and down-regulation of surface antigens, LPO that may lead to protein crosslinking, and damage to multidrug transporters (Dougherty et al., 1998).
Reports that PDT could rapidly induce apoptosis, both in vitro and in vivo, have provided insight into the nature of the photokilling mechanisms. Insight into the mechanism of apoptosis after PDT has perhaps been provided by reports that indicate an association between mitochondrial photodamage and apoptotic responses. Recent studies performed by our group showed that the Bchl based photosensitizers induce the activation of the apoptotic pathway. However, apoptosis is probably not the cause for cell death, since inhibiting the apoptotic pathways did not rescue the cells (Mazor et al. 2003, unpublished).
Reference is made to the following patents and patent applications of the applicants of the present application, the contents of all these patents and patent applications being hereby incorporated by reference in their entirety as if fully disclosed herein: U.S. Pat. Nos. 5,726,169, 5,650,292, 5,955,585, 6,147,195, 6,740,637, 6,333,319, 6,569,846, 7,045,117, DE 41 21 876, EP 1 246 826, WO 2004/045492, WO 2005/120573.